Why AI In Performance Management Falls Short Without Real-Time Signals

Cheryl Johnson is the CTPO of Betterworks, shaping product strategy that redefines performance management practices for global enterprises.

getty

AI has become so embedded in daily workflows that its use is closer to the rule than the exception. The percentage of U.S. employees who reported using AI at work at least a few times a year increased from 40% to 45% between the second and third quarters last year, according to Gallup research.

As it happens, AI is also reshaping how employees gauge their work performance, from brainstorming and writing reviews to creating professional goals.

HR leaders are on board with this development, with 49% of the about 430 HR leaders surveyed by my company considering AI use as the top influencer of employee performance. However, Gartner research found that only 8% of HR leaders feel their managers have the skills to leverage AI effectively.

One issue is that, while AI no doubt saves employees’ valuable time, tracking performance only at the point of review may mean the insights are based on data coming too late for AI to analyze the complete picture.

One way that employees and their managers can benefit from AI-led performance tracking is by continuously incorporating signals. However, this involves a shift in capturing data and understanding what signals are valuable. ​

Why AI Only Shows Up When The Work Is Done

Most of the talk about AI use in performance management entails people writing self-reviews, summarizing feedback and polishing their manager evaluations.

This usage saves time and money, but it also typically occurs after the work is complete. The AI is summarizing performance. It is operating at the moment of reflection, not the moment of performance, which is happening continuously.

This is like using AI to edit a book draft instead of giving the author in-progress tips on narrative flow, possible adjustments to character arcs and other strategic tips. In the workplace, that intelligence is too often applied at the least leveraged moment in the performance cycle.​

Only focusing on performance at a particular point in time can, as a Forrester analyst pointed out, lead to reviews that "are limited, arbitrary and biased." ​

Applying AI To Skills And Performance​

When integrated at various points in time—such as during one-on-ones, feedback sessions and skills development—AI can offer real-time visibility into skills development and surface problems before performance stalls​.​

AI-powered tools can infer skills from contributions over time, with visibility into skill development through goal accomplishment and feedback. What is the employee learning about their skills from peers, from one-on-ones and from managers' weekly meetings?

AI also enables a deeper look into the employee’s context. What is their role and job family? How does performance evidence over time relate to recommendations for skill development? AI can detail how different sources indicate whether the employee has or lacks certain skills.

These insights can help managers identify more objectively where employees need guidance, while also allowing them to spend more time focused on coaching. Likewise, for employees, having feedback that represents the scope of their work can help them better align with expectations, which can in turn lead to improved retention. ​

Implementing A Real-Time Performance Strategy

Of course, as with any new technology, there are tradeoffs.

Organizations often struggle not with AI itself, but with how they apply it, such as by layering it onto fragmented, episodic systems instead of embedding it into the flow of work.

When AI is disconnected from real work signals like feedback from one-on-ones and progress on goals, insights are incomplete. This can reduce confidence in the system while also leaving performance dependent on subjective judgment. ​

There is also a risk of over-automation. Optimizing only based on AI outputs can lead managers to defer judgment to AI, rather than using their judgment to fill in the gaps. Without the right guardrails, AI can accelerate existing biases or create confusion about how decisions are made.​

To succeed, organizations must anchor performance in outcomes—not activity—and embed AI into real-time workflows. Just as important is investing in manager effectiveness and change management, ensuring leaders are equipped to interpret AI-driven insights, have better coaching conversations and build trust with their teams.

Best practices include clearly defining which decisions AI informs versus those that remain human, improving data hygiene and system integration, maintaining transparency into how data is used and creating feedback loops so employees can understand and challenge outputs.

Starting with focused use cases, rather than enterprise-wide rollouts, can also help teams build confidence and adoption over time.

When done right, AI doesn’t replace performance management—it enables a real-time, evidence-based approach that drives better decisions, stronger execution and a clearer connection between people performance and business outcomes.​​

Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?